Summary/Abstract Transient absorption microscopy has enormous potential for providing label-free imaging contrast from endogenous molecules. However, the technique is seriously limited by its requirement for UV/visible-wavelength interactions. At these wavelengths, optical absorption and scattering are high, limiting penetration depth, causing excess sample heating, which in turn limits the amount of power that can be applied and signal-to-noise levels that can be recorded. We propose a new approach to transient absorption that mitigates the disadvantages inherent to resonant interactions, while maintaining the multiphoton advantages that motivated its original development. The new approach will open a pathway for transient absorption imaging deep inside of tissues, while improving image quality and expanding the technique to its full range of endogenous molecular targets for label-free in vivo imaging. Aim 1 will move the probe to near- infrared wavelengths, using transient phase detection with a stable inline interferometry technique. Aim 2 will move the pump to the near-infrared region using two-photon absorption. Together, these approaches will eliminate the need for UV/visible wavelengths (resonant absorption), and thereby mitigate the technical hurdles limiting more widespread use of transient absorption microscopy. This new technique will have broad applicability to advance biomedical research, with potential direct clinical applications, for example basic tumor biology and study of microenvironment and chemoresistance, diagnosis and quantitative follow-up of mitochondrial diseases, and staging and surgical planning for melanoma. The off-resonant transient absorption microscopy to be developed in this R21 will open new vistas in molecular imaging targets with high spatial resolution deep inside of tissues.